def _guessYelpId(placeName, lat, lon):
safePlaceId = hashlib.md5(placeName).hexdigest()
cachedId = db.child(eventsTable).child("cache/" + safePlaceId).get().val()
if cachedId:
return cachedId
opts = {
# 'term': placeName, # Yelp does a bad job with term searching
'limit': 20,
#'radius_filter': 1000,
#'sort_by': 'distance',
'sort': 1,
}
r = yelpClient.search_by_coordinates(lat, lon, **opts)
if len(r.businesses) > 0:
location = (lat, lon)
businessesWithCoords = filter(
lambda b:
(b.location is not None) and (b.location.coordinate is not None),
r.businesses)
biz = min(businessesWithCoords, key=lambda b:
geo.distance(location,
(b.location.coordinate.latitude, b.location.coordinate.longitude))
)
log.debug("%s --> %s" % (placeName, biz.name))
researchVenue(biz)
# Add bizId to cache
record = { "cache/" + safePlaceId: str(biz.id) }
db.child(eventsTable).update(record)
return biz.id
else:
log.info("Can't find %s" % placeName)
return None
def get_place_ids_in_radius(center, radius_km, cached_locations_table=None):
"""Returns all place_ids in the DB within a radius from the given center.
:param center: is (lat, lng)
:param cached_locations_table: a direct pull of the geofire `.../location/` table; use this to avoid downloading the table again.
:return: A set of place ids within the given radius.
"""
if cached_locations_table is None:
location_table = _get_locations_table()
else:
location_table = cached_locations_table
place_ids_in_radius = set()
for place_id, vals in location_table.items():
lat, lng = vals['l']
coord = (lat, lng)
if geofire.distance(center, coord) <= radius_km:
place_ids_in_radius.add(place_id)
return place_ids_in_radius
def findSearchRecord(center, radius=1000):
import app.geofire as geo
import time
queries = geo.geohashQueries(center, radius)
now = time.time()
for query in queries:
results = db.child(searchesTable).order_by_key().start_at(query[0]).end_at(query[1]).get()
for result in results.each():
record = result.val()
if record.get("time", 0) + searchCacheExpiry < now:
db.child(searchesTable).child(result.key()).remove()
continue
# double check that we're within distance
circleDistance = geo.distance(center, record["l"]) * 1000
# 1000 m in 1 km (geo.distance is in km, searchCacheRadius is in m)
if circleDistance < searchCacheRadius:
return record
log.info("Circle distance is " + str(circleDistance))
venueData = venueTableComplete["details"]
cacheData = venueTableComplete["cache"]
import json
stats = {
"website": 0,
"TOTAL": 0,
}
factualStats = {}
for venue in venueData.values():
yelpID = venue["id"]
coord = venue["coordinates"]
coord = (coord["lat"], coord["lng"])
if geo.distance(coord, focus) > 500: # 500 km
continue
identifiers = None
try:
identifiers = cacheData[yelpID]["identifiers"]
except KeyError:
identifiers = dict()
for p in identifiers.keys():
factualStats[p] = factualStats.get(p, 0) + 1
print("%.8f, %.8f" % coord)
providers = venue.get("providers", {}).keys()
for p in providers:
stats[p] = stats.get(p, 0) + 1
# Also: this does not attempt to wrap around, or top out at the top or bottom of the globe.
# i.e. this does not generalize to any bounded box on the planet.
def grid_points(north_lat, west_lng, south_lat, east_lng, grid_size_m=100.0):
lat, lng = (south_lat, west_lng)
while lat < north_lat:
lng = west_lng
while lng < east_lng:
yield (lat, lng)
lng += geo.metersToLongitudeDegrees(grid_size_m, lat)
lat_delta = grid_size_m / geo.g_METERS_PER_DEGREE_LATITUDE
lat += lat_delta
grid = grid_points(north_lat, west_lng, south_lat, east_lng, grid_size_m)
grid = [focus] + sorted(grid, key=lambda pt: geo.distance(pt, focus))
office_locations = [
(50.815078, -0.137089), # brighton
(51.385114, -0.008778), # croydon
# (-15.5095764,167.1841149), # vanuatu, lol not really.
(43.6472912, -79.3966112), # toronto
(37.7895639, -122.3911524), # san francisco
(49.2824693, -123.1113848), # vancouver
(52.5417121, 13.3869854), # berlin
(48.8721423, 2.3389602), # paris
(51.5045923, -0.0992805), # london
(45.5234539, -122.6848713), # portland
(37.387319, -122.0622035), # mountain view
(35.6652311, 139.725706), # tokyo
]